I suspect that UIP starts with microscopic dense fibrosis with FF generally located in the base of the lower lobe subpleurally (Fig. 7.3a, b). Gradually, this fibrosis extends upward and to the inner lung continuously or discontinuously until HRCT can detect it (Fig. 7.3c, d).

I personally defined the UIP stages as follows: (a) microscopic stage, when macroscopically undetectable; (b) mild macroscopic stage, macroscopically detectable but up to 1 cm in depth from the pleura; and (c) extensive macroscopic stage, more than 1 cm in depth from the pleura by macroscopic examination of resected lung [13–15]. The mild macroscopic stage roughly correlates with the HRCT-detectable stage when there are faint or clear, sufficient reticular shadows and/or honeycombing. When bilateral lower lobe reticular shadows or honeycombing are clearly seen within 1 cm in depth, it is subclinical IPF. The extensive macroscopic stage might be roughly clinical IPF.

It is unclear how many years are required to progress from onset until the mild macroscopic stage, although 10 to 20 years has been suggested by me. The incidence of macroscopically detectable mild- and extensive-staged UIP was more than 20 % in more-than-moderate smokers and 3.5 % in nonsmokers in lungs subjected to lobectomy for lung cancer [14]. The incidence of mild macroscopic UIP increases with age (0 %, <40 years; 3 %, 50–60; 14.1 %, 70–80; 28 %, >80) [13]. It seems that mild macroscopic UIP begins approximately at the age of the 40s. The rate of subclinical interstitial pneumonia (mainly subclinical IPF) was 9.7 % [16] in Japan. In addition, a rate of bilateral lower lobe reticular shadows of 8 % was reported, as determined by HRCT in asymptomatic smokers [17]. I suppose that both variously typed interstitial pneumonias and airspace enlargement with fibrosis (AEF) [14] are included in this 8 %. As for progression, Nagai et al. reported that routine chest X-ray-screened, pathologically proven UIP cases became symptomatic 1000 days later [18]. Fukushima et al. followed up 127 microscopic and macroscopic mild UIP cases for 4 years; 4 % showed acute exacerbation and 6 % showed chronic progression [19]. It is necessary to follow up cases that show HRCT-detected reticular shadows and macroscopically detected UIP lesions found by lobectomy to determine how many years are required for progression to clinical IPF.

DAD superimposed of UIP is clinically called acute exacerbation [20–30], whether UIP is idiopathic or secondary. Liebow might have been the first to report acute exacerbation of IPF [5]. In Japan, idiopathic interstitial pneumonia was divided into the acute form (idiopathic DAD) and the chronic form (IPF) in 1976 following Liebow’s concept because of the common complication of DAD in UIP. Yoshimura et al. reported acute exacerbation of IPF for the first time in 1984 [20]. Thirty-five cases suffered 43 episodes of acute exacerbation, and 97.1 % had died one month after the episode. They proposed diagnostic criteria of acute exacerbation. Kondo and Saiki published the first English paper in 1989 [21]: (a) clinical study (among 155 IPF cases, 89 (57 %) showed acute exacerbation) and (b) pathological study (among 22 SLB-proved IPF cases, four showed acute exacerbation, and pathological examination confirmed hyaline membrane formation (one type of DAD) on UIP in all of these four cases).

Papers on acute exacerbation were mainly reported from the northeast of Asia, and the incidence of acute exacerbation varied, but was around 10 % per year [20, 22–25]. Risk factors were infection, reduction of corticosteroid dose, and various types of surgery, among others [15, 20–22, 25]. The reported incidences of acute exacerbation of clinical and subclinical IPF following lobectomy for lung cancer are 15 % (collective data of nine papers) [26] and 15.8 % (11 papers) [27]. The incidences of acute exacerbation of microscopic and macroscopic mild and extensive UIP cases just after lobectomy were 1.6, 6, and 10.6 % [15]. The incidence of acute exacerbation following SLB was reported to be low in Europe and America [28, 29].

Generally, the concept of acute exacerbation has been used for clinical IPF. We reported acute exacerbation of macroscopic mild UIP cases following lobectomy in 2001 [13] and again in 2005 [15], in the same year as the report of Chida et al. [30], as well as a case report in 2011 [31]. We also reported acute exacerbation of the microscopic-staged cases [15, 32].

The mortality rate of acute exacerbation varies [20–22, 25, 27], from 20 to 100 %. Kondoh et al. reported three SLB cases of acute exacerbation without death [33].

The pathological risk factors of acute exacerbation have not been well clarified. Fukushima et al. compared a thick-walled, small-sized honeycombing (4.3 ± 0.5 mm in diameter) cases and a thin-walled, variously sized honeycombing (10.0 ± 1.6 mm in diameter) cases. The former cases showed a higher incidence of acute exacerbation, as well as higher rates of women, nonsmokers, and those with a lower smoking index [34]. We also confirmed that the former cases showed a higher incidence of acute exacerbation, in addition to extent by another trial [15]. Histological factors related to acute exacerbation was the presence of active inflammatory changes: (a) degree of interstitial inflammation continuous with dense fibrosis, (b) amount of granulation tissue in and next to dense fibrosis (FF), (c) quantification of interstitial inflammatory change with granulation tissue apart from dense fibrosis, and (d) quantification of acute interstitial inflammatory change with fibrin exudation seen in the lobectomy UIP area [35]. In addition, a microscopic organized membranous organization pattern of DAD, which will be described later, was seen only in the acute exacerbation cases [35].

According to my experience, two histological patterns of DAD are observed. One is extensive epithelial erosion and hyaline membrane formation with subsequent membranous or ring organization (hyaline membrane was replaced by membranous granulation tissue due to activation of fibroblast) over the orifice ring of the alveolar wall (Fig. 7.4), and the other is epithelial erosion and massive exudation of fibrin following subsequent obstructive and incorporated-type intraluminal organization (Fig. 7.5) [36]. I named these as follows: (a) membranous organization pattern and (b) luminar organization pattern (somewhat resembling OP) in 1992 and 1994 [37, 38] (Figure 7.6 with permission from Kokyu and Modern Physician). Frequently, both patterns and mixed pathology are seen in autopsy lungs, especially cases without a fulminant course, and the luminar organization pattern is older than the membranous organization pattern. DAD pathology as reported by Kondoh et al. was a luminar organization pattern [33]. Mandal et al. reported that DAD with organizing pneumonia showed greater survival than DAD without it (67 % versus 33 %) [39]. A luminal organization pattern and DAD with organizing pneumonia might involve nearly the same pathology. Parambil et al. reported seven cases showing DAD (six cases) and OP (one case) with 86 % mortality [40]. Churg et al. reported 12 cases showing DAD (four cases), OP (five cases), and giant fibroblastic foci (three cases), ten cases of which survived due to therapy [41].

Before writing about pathology of acute exacerbation, I wish to discuss acute lung injury patterns (ALI/P). ALI/P was proposed by Katzenstein including DAD and BOOP in 1997 [42]. As far as my understanding from English literatures, the distinction between organizing DAD (luminal organization pattern) and OP is not so clear to me. The intraluminal organization seen in DAD is obstructive and incorporated mural pattern with an unclear lung structure, and the intraluminal organization in OP is mainly of the polypoid type with a preserved lung structure. Histologically, differential diagnosis between DAD and OP is clear, but there is a transitional-typed organization between DAD and OP. I named this ALI/P not otherwise specified (NOS). Katou et al. also reported this type of acute exacerbation [43]. Table 7.1 presents the histological features and prognosis of the two types of DAD and ALI/P NOS. Nowadays, above three comprise pathological features of acute exacerbation.

As stated before, the mortality rate of acute exacerbation ranges from 0 to 100 %. This difference in the mortality rate might mainly depend on different histology, in addition to recent advanced therapy, and other factors.

As stated before, Flaherty et al. reported that, among 109 cases (either UIP or NSIP), biopsied multiple lobes showed UIP-UIP in 47 %, UIP-NSIP in 26 %, and NSIP-NSIP in 27 % [47]. Monaghan et al. reported that among 64 previously diagnosed IPF cases, 12.5 % showed UIP-NSIP, 39.1 % showed UIP-UIP, and 48.4 % showed NSIP-NSIP [48]. I wonder whether the UIP-NSIP group showed pure UIP-pure NSIP without mixed features in one slide. In both studies, the age of UIP-NSIP cases was intermediate between those of UIP-UIP cases and NSIP-NSIP cases. Our SIP + UIP cases are old (not statistically significant) compared with SIP only. I suspect that, in the UIP-NSIP group, UIP might have a macroscopic mild extent and NSIP might be complicated, which could explain why the patients are younger than those with UIP-UIP and show symptoms. Katzenstein et al. also stated that focal NSIP was frequently seen in UIP [11, 12]. The data indicate that mixed histological patterns frequently exist, even in one disease, and the complication of NSIP can be understood as chronic exacerbation of UIP/stepwise progression of it or subacute exacerbation.

Treatment also affects histological features. Matsushima reported that SLB-proved DIP following treatment showed NSIP by later lobectomy for lung cancer [55]. Generally, at an advanced stage, each pathological pattern loses its characteristic or specific features and becomes nonspecific with/without therapy.

In order to make a pathological diagnosis, clinicians have to avoid far-advanced areas and should choose mildly to moderately affected areas, including normal-looking lung or early-stage areas. If possible, clinicians should select two portions showing different features on HRCT. In addition, the tip of the middle lobe or lingula should be avoided because of many nonspecific changes. Afterward, SLB staple should be removed and formalin should be injected, until the pleural surface becomes smooth, through the staple-removed area, in order to avoid injection of formalin into a fibrotic portion and interlobular septum.